ESQL - TO_DENSE_VECTOR function and additional casting

x-pack/plugin/esql/qa/testFixtures/src/main/resources/dense_vector.csv-spec

@@ -45,3 +45,59 @@ id:l | new_vector:dense_vector
 2    | [9.0, 8.0, 7.0]
 3    | [0.054, 0.032, 0.012]
 ;
+
+convertIntsToDenseVector
+required_capability: dense_vector_field_type
+required_capability: to_dense_vector_function
+
+// tag::to_dense_vector-ints[]
+row ints = [1, 2, 3]
+| eval vector = to_dense_vector(ints)
+| keep vector
+// end::to_dense_vector-ints[]
+;
+
+// tag::to_dense_vector-ints-result[]
+vector:dense_vector
+[1.0, 2.0, 3.0]
+// end::to_dense_vector-ints-result[]
+;
+
+convertLongsToDenseVector
+required_capability: dense_vector_field_type
+required_capability: to_dense_vector_function
+
+row longs = [5013792, 2147483647, 501379200000]
+| eval vector = to_dense_vector(longs)
+| keep vector
+;
+
+vector:dense_vector
+[5013792.0, 2147483647.0, 501379200000.0]
+;
+
+convertDoublesToDenseVector
+required_capability: dense_vector_field_type
+required_capability: to_dense_vector_function
+
+row doubles = [123.4, 567.8, 901.2]
+| eval vector = to_dense_vector(doubles)
+| keep vector
+;
+
+vector:dense_vector
+[123.4, 567.8, 901.2]
+;
+
+convertHexStringToDenseVector
+required_capability: dense_vector_field_type
+required_capability: to_dense_vector_function
+
+row hex_str = "0102030405060708090a0b0c0d0e0f"
+| eval vector = to_dense_vector(hex_str)
+| keep vector
+;
+
+vector:dense_vector
+ [1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0]
+;

x-pack/plugin/esql/qa/testFixtures/src/main/resources/knn-function.csv-spec

@@ -1,7 +1,3 @@
-# TODO Most tests explicitly set k. Until knn function uses LIMIT as k, we need to explicitly set it to all values
-# in the dataset to avoid test failures due to docs allocation in different shards, which can impact results for a 
-# top-n query at the shard level 
-
 knnSearch
 required_capability: knn_function_v5
 
@@ -410,3 +406,52 @@ host:keyword | semantic_text_dense_field:text
 "host1"      | live long and prosper
 ;
 
+
+knnWithCasting
+required_capability: knn_function_v5
+required_capability: to_dense_vector_function
+
+from colors metadata _score
+| eval query = to_dense_vector([0, 120, 0])
+| where knn(rgb_vector, query) 
+| sort _score desc, color asc
+| keep color, rgb_vector
+| limit 10
+;
+
+color:text | rgb_vector:dense_vector
+green      | [0.0, 128.0, 0.0]
+black      | [0.0, 0.0, 0.0]
+olive      | [128.0, 128.0, 0.0]
+teal       | [0.0, 128.0, 128.0]
+lime       | [0.0, 255.0, 0.0]
+sienna     | [160.0, 82.0, 45.0]
+maroon     | [128.0, 0.0, 0.0]
+navy       | [0.0, 0.0, 128.0]
+gray       | [128.0, 128.0, 128.0]
+chartreuse | [127.0, 255.0, 0.0]
+;
+
+knnWithHexStringCasting
+required_capability: knn_function_v5
+required_capability: to_dense_vector_function
+
+from colors metadata _score
+| where knn(rgb_vector, "007800") 
+| sort _score desc, color asc
+| keep color, rgb_vector
+| limit 10
+;
+
+color:text | rgb_vector:dense_vector
+green      | [0.0, 128.0, 0.0]
+black      | [0.0, 0.0, 0.0]
+olive      | [128.0, 128.0, 0.0]
+teal       | [0.0, 128.0, 128.0]
+lime       | [0.0, 255.0, 0.0]
+sienna     | [160.0, 82.0, 45.0]
+maroon     | [128.0, 0.0, 0.0]
+navy       | [0.0, 0.0, 128.0]
+gray       | [128.0, 128.0, 128.0]
+chartreuse | [127.0, 255.0, 0.0]
+;

x-pack/plugin/esql/qa/testFixtures/src/main/resources/vector-cosine-similarity.csv-spec

@@ -90,17 +90,40 @@ total_null:long
 59
 ;
 
-# TODO Need to implement a conversion function to convert a non-foldable row to a dense_vector
-similarityWithRow-Ignore
+similarityWithRow
 required_capability: cosine_vector_similarity_function
+required_capability: to_dense_vector_function
 
-row vector = [1, 2, 3] 
+row vector = to_dense_vector([1, 2, 3]) 
 | eval similarity = round(v_cosine(vector, [0, 1, 2]), 3) 
+;
+
+vector: dense_vector | similarity:double
+[1.0, 2.0, 3.0]      | 0.978  
+;
+
+similarityWithVectorField
+required_capability: cosine_vector_similarity_function
+required_capability: to_dense_vector_function
+
+from colors
+| where color != "black" 
+| eval query = to_dense_vector([0, 255, 255])
+| eval similarity = v_cosine(rgb_vector, query) 
 | sort similarity desc, color asc 
 | limit 10
 | keep color, similarity
 ;
 
-similarity:double
-0.978  
+color:text     | similarity:double
+cyan           | 1.0
+teal           | 1.0
+turquoise      | 0.9890533685684204
+aqua marine    | 0.964962363243103
+azure          | 0.916246771812439
+lavender       | 0.9136701822280884
+mint cream     | 0.9122757911682129
+honeydew       | 0.9122424125671387
+gainsboro      | 0.9082483053207397
+gray           | 0.9082483053207397  
 ;

x-pack/plugin/esql/qa/testFixtures/src/main/resources/vector-dot-product.csv-spec

@@ -88,17 +88,39 @@ total_null:long
 ;
 
 
-# TODO Need to implement a conversion function to convert a non-foldable row to a dense_vector
-similarityWithRow-Ignore
+similarityWithRow
 required_capability: dot_product_vector_similarity_function
+required_capability: to_dense_vector_function
 
-row vector = [1, 2, 3] 
+row vector = to_dense_vector([1, 2, 3])
 | eval similarity = round(v_dot_product(vector, [0, 1, 2]), 3) 
+;
+
+vector: dense_vector | similarity:double
+[1.0, 2.0, 3.0]      | 4.5
+;
+
+similarityWithVectorField
+required_capability: dot_product_vector_similarity_function
+required_capability: to_dense_vector_function
+
+from colors
+| eval query = to_dense_vector([0, 255, 255])
+| eval similarity = v_dot_product(rgb_vector, query) 
 | sort similarity desc, color asc 
 | limit 10
 | keep color, similarity
 ;
 
-similarity:double
-0.978  
+color:text | similarity:double
+azure      | 65025.5
+cyan       | 65025.5
+white      | 65025.5
+mint cream | 64388.0
+snow       | 63750.5
+honeydew   | 63113.0
+ivory      | 63113.0
+sea shell  | 61583.0
+lavender   | 61200.5
+old lace   | 60563.0
 ;

x-pack/plugin/esql/qa/testFixtures/src/main/resources/vector-hamming.csv-spec

@@ -87,17 +87,39 @@ total_null:long
 59
 ;
 
-# TODO Need to implement a conversion function to convert a non-foldable row to a dense_vector
-similarityWithRow-Ignore
+similarityWithRow
 required_capability: hamming_vector_similarity_function
+required_capability: to_dense_vector_function
 
-row vector = [1, 2, 3] 
+row vector = to_dense_vector([1, 2, 3])
 | eval similarity = round(v_hamming(vector, [0, 1, 2]), 3) 
+;
+
+vector: dense_vector | similarity:double
+[1.0, 2.0, 3.0]      | 4.0
+;
+
+similarityWithVectorField
+required_capability: hamming_vector_similarity_function
+required_capability: to_dense_vector_function
+
+from colors
+| eval query = to_dense_vector([0, 255, 255])
+| eval similarity = v_hamming(rgb_vector, query) 
 | sort similarity desc, color asc 
 | limit 10
 | keep color, similarity
 ;
-
-similarity:double
-0.978  
+
+color:text | similarity:double
+red        | 24.0
+orange     | 20.0
+gold       | 18.0
+indigo     | 18.0
+bisque     | 17.0
+maroon     | 17.0
+pink       | 17.0
+salmon     | 17.0
+black      | 16.0
+firebrick  | 16.0
 ;

x-pack/plugin/esql/qa/testFixtures/src/main/resources/vector-l1-norm.csv-spec

@@ -87,17 +87,39 @@ total_null:long
 59
 ;
 
-# TODO Need to implement a conversion function to convert a non-foldable row to a dense_vector
-similarityWithRow-Ignore
+similarityWithRow
 required_capability: l1_norm_vector_similarity_function
+required_capability: to_dense_vector_function
 
-row vector = [1, 2, 3] 
+row vector = to_dense_vector([1, 2, 3]) 
 | eval similarity = round(v_l1_norm(vector, [0, 1, 2]), 3) 
+;
+
+vector: dense_vector | similarity:double
+[1.0, 2.0, 3.0]      | 3.0
+;
+
+similarityWithVectorField
+required_capability: l1_norm_vector_similarity_function
+required_capability: to_dense_vector_function
+
+from colors
+| eval query = to_dense_vector([0, 255, 255])
+| eval similarity = v_l1_norm(rgb_vector, query) 
 | sort similarity desc, color asc 
 | limit 10
 | keep color, similarity
 ;
-
-similarity:double
-0.978  
+
+color:text | similarity:double
+red        | 765.0
+crimson    | 650.0
+maroon     | 638.0
+firebrick  | 620.0
+orange     | 600.0
+tomato     | 595.0
+brown      | 591.0
+chocolate  | 585.0
+coral      | 558.0
+gold       | 550.0
 ;